KT 기술혁신부문

K Intelligence의 K Model 중 KT가 자체 기술로 개발한 ‘믿:음 2.0'에 대한 Technical Report(이하 기술 보고서)를 발간했습니다.

본 기술 보고서는 믿:음 2.0의 주요 특징, 한국적 AI 개발을 위한 고품질 학습데이터 확보 방법 및 사전/사후 학습 방식, 그리고 성능 및 안정성 등의 벤치마크 결과를 소개하는 자료입니다.

공개된 믿:음 2.0 모델과 기술 보고서가 AI를 활용한 기업의 기술 혁신과 개발자 생태계의 성장에 도움이 되길 바랍니다.

Abstract

We introduce Mi:dm 2.0, a bilingual large language model LLM) specifically engineered to advance KOREACENTRIC AI. This model goes beyond Korean text processing by integrating the values, reasoning patterns, and commonsense knowledge inherent to Korean society, enabling nuanced understanding of cultural contexts, emotional subtleties, and realworld scenarios to generate reliable, culturally appropriate responses. To address limitations of existing LLMs—often caused by insufficient or low-quality Korean data and lack of cultural alignment—Mi:dm 2.0 emphasizes robust data quality through a comprehensive pipeline that includes proprietary data cleansing, high-quality synthetic data generation, strategic data mixing with curriculum learning, and a custom Korean-optimized tokenizer to improve efficiency and coverage. To realize this vision, we offer two complementary configurations: Mi:dm 2.0 Base 11.5B parameters), built with a Depth-up Scaling strategy for general-purpose use, and Mi:dm 2.0 Mini 2.3B parameters), optimized for resource-constrained environments and specialized tasks. Mi:dm 2.0 achieves state-of-the-art performance in Korean-specific benchmarks, with top-tier zero-shot results on KMMLU and strong results in internal evaluations across language, humanities, and social science tasks. The Mi:dm 2.0 lineup is released under the MIT license supporting extensive research and commercial use. By offering these accessible and high-performance Korea-centric LLMs, KT aims to accelerate AI adoption across Korean industries, public services, and education, while strengthening the Korean AI developer community and laying the groundwork for the broader vision of Kintelligence. Our models are available via K intelligence Mi:dm.
For technical inquiries, please contact midm-llm@kt.com